Three of Canada's largest regulated institutions just decided that going it alone on AI infrastructure isn't worth the cost. Scotiabank, Sun Life, and Telus announced Tuesday they're launching the AI Consortium alongside technology firm Lightworks, a joint model designed to build and govern critical AI control infrastructure that each company would otherwise have to develop independently.
The consortium's premise is straightforward. Large, regulated institutions in banking, insurance, and telecommunications all face nearly identical challenges when implementing AI at scale, from integrating diverse platforms and standards to maintaining governance and operational control. Rather than each company solving that problem alone, the Consortium pools engineering talent and research to jointly build mission-critical AI systems and shared intellectual property.
Why Regulated Industries Are Teaming Up
This is a notable departure from how most companies approach AI competitively. Banks, insurers, and telecoms typically guard their technology investments closely, especially anything touching data governance or compliance. But AI infrastructure has become expensive and complex enough that even a bank with Scotiabank's resources, roughly $1.5 trillion in assets, sees an advantage in sharing the engineering lift with peers rather than duplicating it internally.
Scotiabank has been moving aggressively on AI independently as well, having launched Scotia Intelligence earlier this year, a unified enterprise AI platform that now handles more than 40% of client queries and processes 90% of commercial emails according to the bank's own data. The Consortium builds on that momentum, giving Scotiabank a way to co-develop foundational AI control systems while keeping the resulting Canadian-owned intellectual property under member control.
Why This Matters for Business
In my four years advising executives on AI adoption, one theme comes up constantly with regulated industries. The technology itself often isn't the hard part. Governance, oversight, and maintaining operational control as AI scales across an organization is where most companies get stuck. That's exactly the problem the Consortium is designed to solve collectively rather than company by company.
This model is worth watching regardless of your industry. If you operate in a regulated space, whether that's healthcare, finance, or telecommunications, the calculus around building AI infrastructure alone versus pooling resources with trusted peers is shifting. Consortium-style approaches let companies share the enormous upfront engineering cost of AI governance while still retaining control over the resulting IP, a middle path between building everything in-house and relying entirely on third-party vendors.
What to Watch
Canada's push toward Canadian-owned AI infrastructure is part of a broader sovereignty conversation playing out globally, from Abu Dhabi's sovereign AI funds to domestic data centre buildouts across Europe. Expect more regulated industries to explore similar consortium models over the next year as the cost and complexity of building AI control infrastructure alone becomes harder to justify, especially for mid-sized institutions without the balance sheet of a major bank.
